import numpy as np

import pandas as pd

import matplotlib.pyplot as plt

import seaborn as sns

covid_confirmed = pd.read_csv('https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_confirmed_global.csv')

covid_deaths = pd.read_csv('https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_deaths_global.csv')

covid_recovered = pd.read_csv('https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_recovered_global.csv')

display(covid_confirmed)

covid_confirmed['Province/State'].unique()

# Since column Province/State is not all NAN, we will be keeping it

​

covid_confirmed.rename(columns = {'Province/State':'State', 'Country/Region':'Country'}, inplace=True)

covid_deaths.rename(columns = {'Province/State':'State', 'Country/Region':'Country'}, inplace=True)

covid_recovered.rename(columns = {'Province/State':'State', 'Country/Region':'Country'}, inplace=True)

covid_confirmed.iloc[:, 4:].sum()

# Arranging by top most confirmed cases

# possible answer

covid_confirmed.sort_values('12/22/20', ascending=False).head(10).loc[:, ['Country', '12/22/20']]

# OR

covid_confirmed.groupby(by='Country').sum().sort_values(by=['12/22/20'], ascending=False).head(10).loc[:, [ '12/22/20']]

​

covid_df = covid_confirmed.drop(['Lat', 'Long', 'State'], axis=1)

covid_df = covid_df.groupby(['Country']).sum()

covid_df.sort_values(by=covid_df.columns[-1], ascending=False).head(10)

# covid_df.sort_values(by=covid_df.columns[-1], ascending=False).head(10).plot()

# plt.show()

covid_df.sort_values(by=covid_df.columns[-1], ascending=False).head(10).transpose()

covid_df.sort_values(by=covid_df.columns[-1], ascending=False).head(10).transpose().plot(figsize=(15, 10))

# !pip install cufflinks

from plotly.offline import iplot, init_notebook_mode

import cufflinks

cufflinks.go_offline(connected=False)

init_notebook_mode(connected=False)

covid_df.sort_values(by=covid_df.columns[-1], ascending=False).head(10).transpose().iplot()

covid_deaths.groupby(['Country']).sum().sort_values(by=covid_deaths.columns[-1], ascending=False).head(10)

covid_deaths.groupby(['Country']).sum().sort_values(by=covid_deaths.columns[-1], ascending=False).head(10).transpose().iplot()

covid_recovered.groupby(['Country']).sum().sort_values(by=covid_recovered.columns[-1], ascending=False).head(10)

covid_recovered.groupby(['Country']).sum().sort_values(by=covid_recovered.columns[-1], ascending=False).head(10).transpose().iplot()

# s0= covid_df.loc[covid_df.index=='US'].iteritems()

# last=0

# daily_new_us=list()

# for tup in s0:

#     today=tup[1]['US']

#     change= today-last

#     daily_new_us.append(change)

#     last=tup[1]['US']

# daily_new_us=pd.Series(daily_new_us)

# daily_new_us.iplot()

​

covid_df.loc['US'].diff().iplot(title='Daily confirmed cases in US', color='orange')

confirmed_world = covid_confirmed.drop(['State', 'Lat', 'Long'], axis=1).groupby(['Country']).sum()

deaths_world = covid_deaths.drop(['State', 'Lat', 'Long'], axis=1).groupby(['Country']).sum()

recovered_world = covid_recovered.drop(['State', 'Lat', 'Long'], axis=1).groupby(['Country']).sum()

covid_world = pd.DataFrame(confirmed_world.sum(), columns=['Confirmed'])

covid_world['Deaths'] = deaths_world.sum()

covid_world['Recovered'] = recovered_world.sum()

covid_world

covid_world['Active'] = covid_world['Confirmed'] - covid_world['Deaths'] - covid_world['Recovered']

covid_world['Fatality Rate'] = (covid_world['Deaths']/ covid_world['Confirmed']) * 100

covid_world['Recovery Rate'] = (covid_world['Recovered']/ covid_world['Confirmed']) * 100

# covid_world.style.background_gradient(cmap='Blues', subset=['Confirmed'])\

#                 .background_gradient(cmap='Reds', subset=['Deaths'])\

#                 .background_gradient(cmap='Greens', subset=['Recovered'])

covid_world.iplot()

covid_world.iloc[:, -2].iplot(title='World Covid Fatality Rate')

covid_world.iloc[:, -1].iplot(title='World Covid Recovering Rate')

# !pip install folium

# world_map = folium.Map(location=[10, 10], zoom_start=2, control_scale=True)

# world_map

# Remove nan in Lat colmun

covid_confirmed.drop(covid_confirmed[covid_confirmed['Lat'].isnull()].index, inplace=True) 

covid_confirmed.drop(covid_confirmed[covid_confirmed['Lat'].isnull()].index, inplace=True) 

covid_map = folium.Map(location = [15, 0], zoom_start = 2, control_scale= True, max_zoom= 8 , min_zoom= 1, 

                      tiles = 'CartoDB dark_matter')

for i in range( 0, len(covid_confirmed)):

    folium.Circle(location = [covid_confirmed.iloc[i]['Lat'], covid_confirmed.iloc[i]['Long']], 

                 radius = (int(np.log(covid_confirmed.iloc[i, -1] + 1.00001))+0.1)*15000,    

            tooltip = "<h5 style = 'text-align:center;font-weight: bold'>" + covid_confirmed.iloc[i]['Country']+ "</h5>"+

                  "<h6>Confirmed: "+str(covid_confirmed.iloc[i,-1])+"</h6>"

                  "<h6>Deaths: "+str(covid_deaths.iloc[i,-1])+"</h6>",

                 color = 'red', fill = True).add_to(covid_map)

covid_map

​

# covid19_world = covid_confirmed[['Country', 'Lat', 'Long']]

# covid19_world['Confirmed'] = covid_confirmed.iloc[:, -1]

# covid19_world['Deaths'] = covid_deaths.iloc[:, -1]

# covid19_world['Recovered'] = covid_recovered.iloc[:, -1]

# covid19_world

​